G protein-coupled receptor kinase 5 mediates Tazarotene-induced gene 1-induced growth suppression of human colon cancer cells

Constitutive expression vectors encoding myc and his-tagged TIG1A (pTIG1A-myc) and TIG1B (pTIG1B-myc) fusion proteins were constructed. The TIG1A and TIG1B cDNA fragments were amplified from human SC-M1 gastric cancer cells, obtained from Dr. C.-L. Meng (National Defense Medical Center, Taipei, Taiwan), using TIG1A-specific primers (sense, 5'-CTCTGAATTCCTGCGTCCATGCAGCCC-3' and antisense, 5'-GCGCGGATCCGAAATTACTAAGCTCTG-3') and TIG1B-specific primers (sense, 5'-CTCTGAATTCCTGCGTCCATGCAGCCC-3' and antisense, 5'-CGGACGGATCCGGTTTTTCTTACCCAC-3'). Amplified cDNA fragments were digested with EcoRI and BamHI, and then subcloned in-frame into the multicloning site of the pcDNA3.1-myc-his A expression vector (Invitrogen, Carlsbad, CA, USA). To generate inducible TIG1A and TIG1B expression plasmids (pTIG1A-V5 and pTIG1B-V5, respectively), HindIII- and BamHI-digested TIG1 cDNA fragments isolated from pTIG1A-myc or pTIG1B-myc were cloned in-frame into the pGene-V5-his B vector (Invitrogen).

HCT116 colon cancer cells were maintained in growth medium consisting of McCoy's 5A medium (BioWest, Nuaille, France) supplemented with 25 mM HEPES, 26 mM NaHCO₃, 2 mM L-glutamine, 100 units/mL penicillin, 100 μg/mL streptomycin, and 10% fetal bovine serum (FBS; Hyclone, Logan, Utah, USA) at 37°C and 5% CO₂. SW620 cells were maintained in growth medium consisting of Leibovitz's L15 medium (GIBCO BRL, Grand island, NY, USA) supplemented with 25 mM HEPES, 26 mM NaHCO3, 2 mM L-glutamine, 100 units/mL penicillin, 100 μg/mL streptomycin, and 10% FBS at 37°C without CO₂. HT-29, LS174T, and SW480 colon cells were maintained in Dulbecco's Modified Essential Medium (GIBCO BRL) supplemented with 10% FBS. CC-M1, CC-M2, CC-M3, DLD-1, and COLO205 colon cells were maintained in RPMI medium (GIBCO BRL) supplemented with 25 mM HEPES, 26 mM NaHCO₃, 2 mM L-glutamine, 100 units/mL penicillin, 100 μg/mL streptomycin, and 10% FBS at 37°C and 5% CO₂. Colon cancer cells were obtained from Food Industry Research and Development Institute (Sinju, Taiwan).

HCT116 cells plated in 6-cm culture dishes were transfected with expression vectors using liposome-mediated transfection. Briefly, plasmids or small interfering RNAs (siRNAs) and lipofectamine (Invitrogen) were diluted in Opti-MEM medium (GIBCO BRL) and then mixed at room temperature for 15 min. The nucleic acid-lipofectamine complexes were then added to cells for 2.5 h at 37°C. Cells were then refreshed with complete medium for 24 h at 37°C for further analysis.

Inducible TIG1 stable clones were established using the GeneSwitch system (Invitrogen) as previously described with minor modifications [20]. Briefly, HCT116 cells plated in 10-cm dishes were first transfected with the pSwitch expression vector (Invitrogen) that expressed the hybrid regulatory protein using lipofectamine, and then selected in hygromycin (400 μg/mL, Invitrogen)-containing medium. Stable HCT116/Switch clones were selected by limiting dilution and validated by analysing the induced expression of β-galactosidase followed by the transfection with pLacZ/V5 (Invitrogen) and then incubation with 10 nM MFP. HCT116/Switch cells that expressed the MFP-dependent transactivator were selected for subsequent transfection with pTIG1A-V5, pTIG1B-V5, or the control vector (pGene-V5-his B). Stable cells were selected in medium containing hygromycin and zeocin (125 μg/mL, Invitrogen). Inducible expression of TIG1A and TIG1B was validated by immunocytochemistry and Western blot analyses. Stable clones were maintained in McCoy's 5A growth medium supplemented with hygromycin and zeocin. Cells used for further experiments were plated in medium without hygromycin and zeocin. Representative inducible TIG1A (clone 2), TIG1B (clone 8), and control (2G5) stable clones were selected for analysis.

Using similar methods, inducible TIG1A (clone 13), TIG1B (clone 7), and control (clone 1) stable clones were established in SW620 colon cancer cells. Stable cells were maintained in L15 growth medium supplemented with 800 μg/mL hygromycin and 400 μg/mL zeocin. Cells were monitored monthly for mycoplasma contamination using the DNA intercalating dye, 4'6-diamidino-2-phenylindole (Sigma, St. Louis, MO, USA).

HCT116 cells were seeded at 2 × 10⁴ cells per well in 24-well plates overnight and then transfected with constitutive TIG1A or TIG1B expression vector or control vector for 48 h. WST-1 reagent (100 μL Roche Diagnostics, Mannheim, Germany) was then added to the cells for 4 h at 37°C. Absorbance at 450 and 650 nm was assessed. The percentage of proliferative TIG1A or TIG1B transfected cells relative to controls was defined as [(A₄₅₀-A₆₅₀) of TIG1A or TIG1B transfected cells/(A₄₅₀-A₆₅₀) of control transfected cells] × 100%. Alternatively, control or TIG1 stable clones were plated overnight in growth medium free of hygromycin and zeocin. The cells were then incubated in growth medium containing MFP (0.1 - 100 nM) or ethanol (0.1%) vehicle for 48 h. The percentage of MFP-treated proliferative cells relative to controls was defined as [(A₄₅₀-A₆₅₀) of MFP-treated cells/(A₄₅₀-A₆₅₀) of ethanol-treated cells] × 100%. All experiments were performed in triplicate.

HCT116 cells were plated in 6-well dishes and transfected with pTIG1A-myc, pTIG1B-myc expression vector, or control vector for 3 h after which they were incubated in growth medium for 1 day. Cells were then cultured in medium containing 1000 μg/mL G-418 for 8 days with medium changes performed every other day. The number of colonies was determined using crystal violet staining as described previously [21]. All experiments were performed in triplicate.

Inducible TIG1A, TIG1B, and control stable HCT116 cell clones plated overnight in 10-cm dishes in growth medium were stimulated with growth medium containing 5 nM MFP for 24 h. Cells were washed twice with ice-cold PBS (3.2 mM Na₂HPO₄, 0.5 mM KH₂PO₄, 1.3 mM KCl, 135 mM NaCl, pH 7.4), detached by scraping and isolated by centrifugation at 1500 × g for 5 min at 4°C. Three independent samples were collected for microarray analysis.

Total RNA was extracted using the RNeasy system (Qiagen, Valencia, CA, USA). The integrity of the RNA was assessed using a 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). Integrity values exceeded 9.7 in all tested samples. Affymetrix microarray analysis was carried out according to the manufacturer's protocol (Affymetrix, Santa Clara, CA, USA). Briefly, cRNA probes were prepared from 10 μg of total RNA using an Affymetrix One Cycle cDNA synthesis kit (Affymetrix). The biotinylated cRNA was subsequently fragmented and hybridised to each array at 45°C for over 16 h. cRNA was hybridised to the HGU133-2 plus whole genome GeneChip containing 54,675 transcripts. Quality assessment was performed using the GeneChip Operating System 1.1.1 (Affymetrix) using global scaling to a target signal of 500. Median background was 38.0 - 40.7. Median scaling factors were 6.5 - 9.2, and noise values ranged from 1.6 - 1.8. High overall array and RNA quality were confirmed.

Gene expression data underwent secondary analysis using the GeneSpring GX program (Agilent Technologies, Santa Clara, CA, USA). Briefly, gene expression data from the microarray chips were rescaled (after preliminary analysis) for a direct comparison between all chips. Data were filtered based on both Detection Call and Signal Log Ratio for comparisons between groups (control and TIG1A or control and TIG1B). Data declared "Absent" in all samples were filtered out. Samples had to be declared "Present" in at least 2 of 3 chips within a group to be maintained within the set. The log base 2.0 of the fold change was used as an additional filter. Filtered genes identified to be differentially expressed by 2-fold or greater in 2 of 3 chips were analysed for functional gene clusters using GeneSpring. Statistical analyses were performed using one-way ANOVA (P < 0.05) and t-tests. The program was used to determine functional clusters by statistical representation of individual genes in specific categories relative to all genes in the same category on the array. Signal pathways of differentially expressed genes were validated using the Protein Analysis Through Evolution Relationships (PANTHER™) protein classification system (Applied Biosystems, Foster City, CA, USA). Original microarray data have been deposited in the Gene Expression Omnibus data bank (accession number: GSE23882).

Total RNA was extracted using TRIZOL reagent (Invitrogen). cDNA was prepared by incubating a 20 μL mixture containing 5 μg total RNA, 1 unit Moloney murine leukemia virus (MuLV) reverse transcriptase (Invitrogen), 0.5 μg oligo-dT_12-18 in 50 mM Tris-HCl, pH 8.3, 75 mM KCl, 4.25 mM MgCl₂, 0.5 mM dNTP, and 1 unit RNaseout recombinant RNase inhibitor (Invitrogen) at 37°C for 1 h. Real-time RT-PCR was performed using the ABI prism 7900HT sequencer detector system (Applied Biosystems) and Qiagen's Quantitect SYBR Green PCR kit (Qiagen) following the instructions provided. In brief, reaction mixtures (25 μl of total volume) containing 250 ng cDNA, gene-specific forward and reverse primers (1 μM), and 12.5 μL of 2× Quantitect SYBR Green PCR Master mix were cycled as follows: 1 cycle at 95°C for 10 min; 40 cycles at 95°C for 15 seconds; and 60°C for 1 min. Negative controls without the template were run in parallel to assess the overall specificity of the reaction. PCR primers used for amplification included the following: beta-actin (sense, 5'-TCCCTGGAGAAGAGCTACG-3' and antisense, 5'-GTAGTTTCGTGGATGCCACA-3'), TIG1 (sense, 5'-CGTGGTCTTCAGCACAGAGCG-3' and antisense, 5'-CCCAAACGTCCCTCACCTTCC-3'), GRK5 (sense, 5'-GACCACACAGACGACGACTTC-3' and antisense, 5'-CGTTCAGCTCCTTAAAGCATTC-3'), TIG1A (sense, 5'- CCGAGCCAACTTTCCTGCGTCC-3' and antisense, 5'-CCCTGAGGAACCTGCTGGTGACTG-3'), TIG1B (sense, 5'-CGAGCCAACTTTCCTGCGTC-3' and antisense, 5'-CTTGTGGCACAAGTTAATTTTCAGG-3'), and TCF7 (sense, 5'-TCCAGAGCCCCTGGAGGACG-3' and antisense, 5'-GGGCTGATTGGCCTTGTGCA-3'). Relative TIG1A, TIG1B, TCF7, or GRK5 mRNA levels in MFP-induced stable cells after normalisation to the vehicle treated control cells were calculated as follows: 2^{- ΔΔ}C_T = 2/([C_T of TIG1A in MFP-treated cells]- [C_T of actin in MFP-treated cells])- ([C_T of TIG1A in control cells]- [C_T of actin in control cells]). All experiments were performed in triplicate.

Relative expression of TIG1A and TIG1B in normal tissues and colon cancer cell lines was determined using the icycler system (Bio-Rad Laboratories, Hercules, CA, USA). Total RNAs from normal tissues obtained from Clontech Laboratories (Clontech, Palo Alto, CA, USA) and colon cancer cells purified as described above were reverse transcribed. Real time RT-PCR was performed in reaction mixtures (20 μL total volume) containing 100 ng cDNA, gene-specific forward and reverse primers (160 nM), and 10 μL 2× iQ™ SYBR^® Green Supermix (Bio-Rad). The real-time cycler conditions were as follows: 1 cycle at 94°C for 3 min and 40 cycles at 94°C for 30 seconds, 61°C for 30 seconds, and 72°C for 30 seconds. Primers used for amplification were as follows: TIG1A (sense, 5'-CCTGTCGCATTCACTTGGTCTG-3' and antisense, 5'-CGGAGG CTTCTTCTGGTGTCTG-3'), TIG1B (sense, 5'-TCTGAGACTCATCTGGGATTTG-3' and antisense, 5'-CTTTGATTGTAACTCTTGTGGC-3'), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; sense, 5'-TCCATGCCATCACT GCCACCC-3' and antisense, 5'-GACGCCTGCTTCACCACCTTC-3'). Relative TIG1A or TIG1B mRNA levels were calculated using the following equation: 10,000 × 2^-ΔΔC_T = 10,000 × 2/([C_T of TIG1 in tissue A]- [C_T of GAPDH in tissue A])- ([C_T of TIG1 in colon]- [C_T of GAPDH in colon]). All experiments were performed in triplicate.

Total cytosol extracts were prepared in mild lysis buffer (25 mM HEPES, pH 7.5, 150 mM NaCl, 1% Igepal CA-630, 10 mM MgCl₂, 1 mM EDTA, and 10% glycerol) containing protease inhibitors (20 μg/mL aprotinin and 20 μg/mL phenylmethylsulfonyl fluoride) and phosphatase inhibitors (2 mM NaF and 1 mM Na₃VO₄). Proteins (5 to 100 μg) were separated using 10-12% polyacrylamide gels containing sodium dodecyl sulfate and were then transferred to polyvinylidene difluoride membranes. After blocking, membranes were incubated with anti-V5 (Invitrogen), anti-myc (Invitrogen), anti-GRK5 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), or anti-actin (Sigma) antibody for 12 h at 4°C. Membranes were then incubated with horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit antibodies at room temperature for 1 h. Specific protein bands were developed using Amersham ECL (Amersham, Bucks, UK). Relative protein expression levels were quantified by normalising to actin expression levels.

siRNA oligonucleotides were synthesised by Ambion (Austin, TX, USA). Three TIG1 siRNAs targeted to nucleotides 488 to 508 (5'-GUACAAGUUACAUUGAUGGtt-3'), 540 to 560 (5'-GUUGUAAAGCAGGUAAUCCtc-3'), and 596 to 616 (5'-CCAUGAUUAUCAGGUAUGCtg-3') were synthesised based on Genbank accession NM_26963. The TIG1 siRNAs are specific for both TIG1A and TIG1B isoforms. GRK5 siRNAs targeted to nucleotides 452 to 472 (5'-UAACACUCCAGCCCAGGCCtg-3'), 2158 to 2178 (5'-GAAAUUUCUUUGAGAAACCtc-3'), and 2406 to 2426 (5'-GCACAGAAAAUAUACACCCtc-3') were synthesised according to Genbank accession NM_005308. The Silencer^® negative control#1 siRNA (Ambion) was used as a negative control. Inducible TIG1 or control stable cells were cultured for 24 h in 6-well plates and then transfected with 30 nM siRNAs using Lipofectamine according to the instructions provided. After transfection, cells were incubated in the presence or absence of MFP for 48 h, and then analysed for growth.

Numerical data are shown as mean ± standard deviation (SD) of triplicates from each independent biological sample. Cell proliferation, colony formation, and real-time RT-PCR data were analysed using student's t test. P-values less than 0.05 were considered to be statistically significant. Microarray data were analysed by one-way ANOVA using GeneSpring GX software.

Relative expression levels of TIG1A and TIG1B were evaluated in normal human tissues and in colon cancer cell lines using real-time RT-PCR. Both TIG1A and TIG1B isoforms were expressed at high levels in the prostate and colon. However, expression was barely detectable in bone marrow and skeletal muscle (Figure 1A). Compared with normal colon tissue, expression of both TIG1 isoforms was decreased in all colon cancer cell lines analysed (Figure 1B). Seven of the cell lines analysed expressed less than 2 and 11% of normal TIG1A and TIG1B expression levels, respectively.

The activities of TIG1A and TIG1B were measured in HCT116 cells that exhibited high transfection efficiency and were previously shown to have a hypermethylated TIG1 gene. Ectopically expressed recombinant TIG1A and TIG1B protein isoforms with the expected molecular weights of 36.7 and 29.2 kDa, respectively, were detected in HCT116 cells after transient transfection for 24 h (Figure 2A). The effects of TIG1A and TIG1B expression on HCT116 cell growth were first analysed using the WST-1 cell proliferation assay. After 48 h, HCT116 cell growth was significantly inhibited by 27 and 25% following transient transfection with constitutive TIG1A and TIG1B expression vectors, respectively (Figure 2B). Similarly, colony formation of HCT116 cells was significantly suppressed by 63.3 and 89.2% upon transient transfection with TIG1A and TIG1B expression vectors, respectively and maintenance in G418-containing media for 8 days (Figure 2C). The effect of TIG1 isoform expression on cell death was also determined by measuring the release of lactate dehydrogenase in transiently transfected HCT116 cells. Significantly increased cell death (10.3 to 24.6%) was detected in TIG1B-transfected cells; however, cell death was weakly, but significantly, repressed by 5.3 to 12.4% in TIG1A-transfected cells (Additional file 1).

As production of stable cells constitutively expressing TIG1A or TIG1B protein was unsuccessful, inducible TIG1A, TIG1B, and control stable cells were established from HCT116 cells using the GeneSwitch system [18] in which target protein expression is induced by exposure to MFP. Exposure to 5 nM for 24 to 48 h profoundly increased expression of TIG1A or TIG1B in respective TIG1A or TIG1B stable cells, but not the control stable cells (Figure 2D). Expression of both TIG1A and TIG1B was induced after exposure to 0.1 - 100 nM MFP for 48 h (data not shown).

To verify that TIG1A and TIG1B expression influenced HCT116 cell growth, proliferation was assessed in stable clones incubated with various concentrations of MFP (Figure 2E). Cell proliferation, as determined by WST-1 cell proliferation assay, was inhibited by MFP (0.1 - 100 nM) in control stable cells with a maximal inhibition of 25.4% by 100 nM MFP. Induction of TIG1A and TIG1B expression by MFP treatment resulted in a concentration-dependent inhibition of cell growth in both TIG1A and TIG1B stable cells. When compared with MFP-treated control stable cells, 29.6 and 22.2% inhibition of cell growth was observed in TIG1A and TIG1B cells, respectively, incubated with 100 nM MFP for 48 h (Figure 2E). The release of lactate dehydrogenase as an indicator of cell cytotoxicity in stable cells exposed to 5 nM MFP for 1 to 2 days was also analysed; however, no increase in cell death was observed (data not shown). As 5 nM MFP efficiently induced expression of recombinant TIG1 isoforms, this concentration was chosen for microarray analysis to avoid the identification of genes differentially regulated by TIG1A or TIG1B in cells exposed to a high concentration of MFP.

To examine the effects of TIG1A and TIG1B on the gene expression profile of HCT116 cells, stable cells were incubated with 5 nM MFP for 24 h after which they were analysed using the human HGU133 Plus 2.0 microarray (Additional file 2). Genes differentially expressed by greater than two-fold (relative to expression in control cells) in TIG1A (Additional file 3) or TIG1B (Additional file 4) stable cells were selected for analysis. Following induction of TIG1A expression, a total of 108 and 21 genes were significantly upregulated or downregulated, respectively (Additional file 3). Similarly, upon induction of TIG1B expression, a total of 41 and 14 genes were upregulated or downregulated, respectively (Additional file 4). A total of 23 of the upregulated genes and 6 of downregulated genes identified were detected upon both TIG1A and TIG1B induction (Figure 3, Additional file 5). Among the 29 genes regulated by both TIG1A and TIG1B isoforms, genes related to the p53 (wild type p53-induced gene 1 [WIG1] and small ubiquitin-like modifier 2 [SUMO2]), Wnt (homeobox B7 [HOXB7]), interleukin (eukaryotic translation elongation factor 1 epsilon 1 [EEF1E1]), chemokine and cytokine (GRK5 and EEF1E1), and G-protein coupled receptor (GCPR; GRK5) signaling pathways were identified.

The Wnt signaling pathway is stimulated by the GPCR ligand, prostaglandin E2, through enhanced nuclear β-catenin localisation and β-catenin/TCF-mediated transactivation [22]. GRK5 is a member of family proteins that attenuate GPCR activities. Therefore, GRK5 and TCF7 were selected for further analysis. Upon exposure to 5 nM MFP for 24 h, TIG1A and TIG1B mRNAs levels were increased by 107.1- and 29.2-fold, respectively (Figure 4A). Compared with control stable cells, basal TIG1A and TIG1B mRNA levels in TIG1A and TIG1B stable cells not exposed to MFP were 166.2- and 37.8-fold higher, respectively. These results suggest the presence of a weak basal activity of the inducible promoter although basal levels of TIG1 fusion proteins were not detected by Western blot analysis (Figure 2D and Figure 4B). Induced expression of TIG1A and TIG1B by MFP treatment resulted in significantly increased GRK5 (24.3- and 45.9-fold, respectively) and TCF7 (16.3- and 35.8-fold, respectively) mRNA levels (Figure 4A). Basal levels of GRK5 and TCF7 mRNA in TIG1A and TIG1B stable cells were similar to those in control cells.

TIG1-mediated regulation of GRK5 and TCF7 protein expression was examined using Western blot analysis. No GRK5 protein was detected in MFP-treated control cells or in TIG1A and TIG1B stable cells in the absence of MFP treatment. Increased GRK5 protein expression (65 kDa) was observed in TIG1A and TIG1B stable cells exposed to 5 nM MFP for 24 to 48 h (Figure 4B). Compared with cells exposed to MFP for 24 h, GRK5 protein levels were increased by 22 and 75% in TIG1A and TIG1B stable cells exposed to MFP for 48 h, respectively. Similarly, basal GRK5 protein was not detected in HCT116 cells (Figure 4C), while GRK5 protein expression was upregulated in HCT116 cells transiently transfected with the constitutive TIG1A expression vector for 24 to 72 h. GRK5 protein levels were increased (relative to 24 h transfected cell levels) by 58 and 179% after transfection for 48 and 72 h, respectively. Regardless of TCF7 mRNA upregulation, neither TIG1A nor TIG1B expression influenced TCF7 protein levels after exposure to MFP for 24 and 48 h (Figure 4B).

To evaluate the role of GRK5 in TIG1-mediated inhibition of HCT116 cell proliferation, RNA interference was employed. TIG1A stable cells were transiently transfected with siRNAs specific for both TIG1 isoforms or GRK5 and then immediately exposed to 5 nM MFP for 24 h. TIG1A protein levels were reduced by 53.7% in cells transfected with TIG1 siRNA, while GRK5 protein levels was reduced by 51.7% in GRK5 siRNA transfected cells, determined by Western blotting analysis (Figure 5A). In control siRNA transfected TIG1A cells, cellular growth was significantly inhibited to 70.2% of control levels upon MFP exposure that induced TIG1A expression. Silencing of TIG1A or GRK5 expression significantly alleviated TIG1A-mediated growth suppression to 80 and 88.2% of control growth, respectively (P < 0.05 and P < 0.001, respectively; Figure 5B). Levels of cell growth in MFP-treated TIG1A cells transfected with GRK5 siRNA were slightly higher, but not significant, than that of cells transfected with TIG1 siRNA.

To verify the cell independent effect of TIG1 isoform expression in colon cancer cells and to evaluate the effect of TIG1 on cells established from metastatic lesions, TIGA, TIG1B, and control stable clones were established from SW620 colon cancer cells derived from a colon cancer lymph node metastasis. Compared with HCT116 cells, relative TIG1A mRNA levels were higher in SW620 cells. However, TIG1B mRNA levels were lower in SW620 cells (Figure 6A). Upon exposure to 5 nM MFP for 24 to 48 h, enhanced GRK5 protein expression was accompanied by induced expression of the TIG1A or TIG1B recombinant protein isoform in the respective stable, but not in control cells (Figure 6B). Growth in TIG1A and TIG1B stable SW620 cells was decreased (compared to growth in control stable cells) by 39 and 39.8%, respectively, upon exposure to 100 nM MFP (Figure 6C). The release of lactate dehydrogenase was determined in stable cells exposed to 5 nM MFP for 1 to 2 days; no increase in cell death was observed (data not shown).